Intelligent Identification Of Blasting Fragmentation And Parameter Dynamic Optimization Of Blasting In Underground Mine

Blasting effect evaluation and parameter optimization of underground mine are always one of the important research directions of mining industry.For a long time,the blasting effect evaluation methods of mining enterprises are backward,and can not be organically combined with parameter optimization,which leads to the solidification of blasting parameters.A mine almost uses the same set of parameters,resulting in the problems of large randomness and poor reliability of blasting effect.In order to obtain stable blasting effect,it is necessary to evaluate the blasting effect quickly and accurately,and optimize the parameters dynamically combined with the blasting effect.In blasting engineering,the boulder rate is one of the important indexes to evaluate the blasting effect.The boulder rate can directly reflect the blasting quality.At present,the statistical method of mine bulk rate is clumsy and greatly affected by human factors.Under the condition of tight production and continuous blasting,it is far from meeting the requirements of efficient,economic and intelligent production.Therefore,based on the engineering background of qianyanshan iron mine and the actual production data,this paper carried out the research on intelligent identification of blasting fragmentation and dynamic optimization of parameters in underground mine.Firstly,Kuz-Ram model is used to predict the fragmentation distribution after blasting.Then,the yolact deep learning model is used for fast statistics of large ore and rock.At the same time,the blasting parameters are dynamically optimized according to different blasting effects,so as to change the problem of invariable blasting parameters.Finally,the dynamic optimization system of blasting parameters in underground mine is developed,which realizes the functions of blasting fragmentation prediction,blasting fragmentation evaluation and dynamic optimization of blasting parameters.It effectively improves the accuracy of blasting parameter design and optimizes the blasting effect.It has important theoretical and practical significance for efficient and intelligent mining of underground mine and improving mine operation efficiency.The specific research contents of this paper are as follows:(1)Prediction of blasting fragmentation in underground mines.Based on the investigation of the mine in front of the mountain,combined with the distribution law of blasting fragmentation,this paper uses Kuz-Ram model to predict the initial parameters of blasting fragmentation.Based on Kuz-Ram model,the quantitative relationship between blasting parameters(such as explosive consumption,hole network parameters,charge parameters,etc.)and blasting fragmentation information is established.In this paper,the quantitative relationship is used to predict blasting fragmentation information,and C#programming language is used to visualize the predicted fragmentation information and facilitate the analysis of blasting parameters and fragmentation information.(2)Evaluation of blasting fragmentation in underground mines.Based on deep learning technology,the yloact model is used to detect large blocks in the image of underground mine.First of all,the image of underground mine pile blasting is obtained through field investigation,and the image is processed by using preprocessing technology.Then,label the ore in the image with the label tool,and finally,train the model.The trained yloact model can better detect the large ore in the blasting pile and get the large ore rate of the blasting pile.The blasting effect is evaluated by combining the perforation cost,explosive consumption,secondary crushing and shovel loading efficiency.The evaluation results are divided into four grades.(3)Optimization of blasting parameters in underground mine.In this paper,the main factors affecting the blasting effect are analyzed systematically.Combined with the quantitative relationship between blasting parameters and blasting fragmentation information established by Kuz-Ram model,the index and weight of parameter optimization are determined.According to the four grades of blasting effect evaluation,the empirical formula is used to optimize the initial blasting parameters to obtain more reasonable blasting parameters,It is convenient for subsequent blasting engineering to obtain better blasting effect.(4)Research and development of intelligent recognition and parameter dynamic optimization system for blasting fragmentation in underground mine.Through the in-depth study of blasting fragmentation prediction,fragmentation evaluation and parameter optimization.This paper uses C#,Python programming language,combined with visual studio 2015 development platform,based on.Net framework 4.0 framework,to complete the main function development of underground mine blasting fragmentation intelligent identification and parameter dynamic optimization system.Its main functions include five modules,namely:registration module,blasting parameter input module,blasting fragmentation prediction module,blasting fragmentation evaluation module,blasting fragmentation evaluation module,blasting fragmentation evaluation module,blasting fragmentation prediction module,blasting fragmentation evaluation module Optimization module of blasting parameters.